Project Description

The research practicum involves on-site experiential learning in a research setting. This setting may be in the private or public sector, it may include such locations as education, governmental, non-governmental, or general research organization. The experience must provide students the opportunity to collect and analyze data, consider ethical implications of research, and draw empirically grounded conclusions.

Purpose:
Carry out exploratory data analysis on a set of random sample data extracted for machine learning.
Universtiy Name: Utica College
Course Name: DSC-680-Z1 Research Practicum
Student Name: Henry J. Hu
Program Director Name: Dr. McCarthy, Michael
Runtime Environment: RStudio
Programming Language: R
Original Data Frame: 12,705,553 international wires belonging to 139 customers from 3 continents for the entire year of 2020.
Last Update: July 21st, 2021

Clearing R Studio Memory Usage

gc()
##           used (Mb) gc trigger (Mb) max used (Mb)
## Ncells  540773 28.9    1234400   66   621331 33.2
## Vcells 1018036  7.8    8388608   64  1601285 12.3
rm(list = ls())

Time Counter Start

start_time <- Sys.time()

Include the knitr package for integration of R code into Markdown

knitr::opts_chunk$set(echo = TRUE)

All the libraries used in this code

library(readr)
library(easypackages)
libraries("caret","caretEnsemble","caTools","class","cluster","data.tree","devtools","doSNOW","dplyr","e1071","factoextra","gbm","FNN","FSelector","ggalt","ggforce","ggfortify","ggplot2","gmodels","klaR","lattice","mlbench","modeest","nnet","neuralnet","outliers","parallel","psych","purrr","readr","rpart","rpart.plot","spatialEco","stats","tidyr","randomForest","ROSE","rsample","ROCR","pROC","glmnet","gridExtra","R6","Epi")

Import data into RStudio

input_data <- read_delim("Final_cleaned_data.txt", ",", escape_double = FALSE, col_types = cols(
              TRANSACTION_ID = col_character(),
              TRANSACTION_TIME = col_datetime(),
              TRXN_MONTH = col_character(),
              CLIENT_ID = col_character(),
              COUNTRY_NAME = col_character(),
              COUNTRY_CODE = col_character(),
              CONTINENT_NAME = col_character(),
              CONTINENT_CODE = col_character(),
              SWIFT_MSG_TYPE = col_character(),
              AVG_TRXN_AMT = col_double(),
              TRANSACTION_AMOUNT = col_double()
              ),
    trim_ws = TRUE)

Sample data for data exploratory analysis

This sample data is for exploratory data analysis only.

# Set random seed
set.seed(42)

# Sample the data
# input_data_4M <- input_data[sample(nrow(input_data), 4000000), ]

# Write data to storage
# write.table(input_data_4M, file="sample_df_4M.txt", append = FALSE, sep = ",", dec = ".", row.names = FALSE, col.names = TRUE)
# write.csv(input_data_4M,"sample_df_4M.txt", row.names = FALSE)

# Sample the data
# input_data_100K <- input_data[sample(nrow(input_data), 100000), ]

# Write data to storage
# write.table(input_data_100K, file="sample_df_100K.txt", append = FALSE, sep = ",", dec = ".", row.names = FALSE, col.names = TRUE)
# write.csv(input_data_100K,"sample_df_100K.txt", row.names = FALSE)

# Load data into data frame
input_data_eda <- read_delim("sample_df_4M.txt", ",", escape_double = FALSE, col_types = cols(
              TRANSACTION_ID = col_character(),
              TRANSACTION_TIME = col_datetime(),
              TRXN_MONTH = col_character(),
              CLIENT_ID = col_character(),
              COUNTRY_NAME = col_character(),
              COUNTRY_CODE = col_character(),
              CONTINENT_NAME = col_character(),
              CONTINENT_CODE = col_character(),
              SWIFT_MSG_TYPE = col_character(),
              AVG_TRXN_AMT = col_double(),
              TRANSACTION_AMOUNT = col_double()
              ),
    trim_ws = TRUE)

Sample data for data for plotting

This sample data is for plotting only.

# Sample the data
# input_data_100K <- input_data[sample(nrow(input_data), 100000), ]

# Write data to storage
# write.table(input_data_100K, file="sample_df_100K.txt", append = FALSE, sep = ",", dec = ".", row.names = FALSE, col.names = TRUE)

# Load data into data frame
input_data_plot <- read_delim("sample_df_100K.txt", ",", escape_double = FALSE, col_types = cols(
              TRANSACTION_ID = col_character(), 
              TRANSACTION_TIME = col_datetime(),
              TRXN_MONTH = col_character(),
              CLIENT_ID = col_character(), 
              COUNTRY_NAME = col_character(), 
              COUNTRY_CODE = col_character(), 
              CONTINENT_NAME = col_character(), 
              CONTINENT_CODE = col_character(), 
              SWIFT_MSG_TYPE = col_character(), 
              AVG_TRXN_AMT = col_double(),
              TRANSACTION_AMOUNT = col_double()   
              ),
    trim_ws = TRUE)

Descriptive Statistics

These descriptive statistics reveal both the central tendency and dispersion tendency of the sample data for machine learning.

Dimension of data frame

dim(input_data_eda)
## [1] 4000000      11

Structure of data frame

str(input_data_eda)
## tibble [4,000,000 x 11] (S3: spec_tbl_df/tbl_df/tbl/data.frame)
##  $ TRANSACTION_ID    : chr [1:4000000] "3174204" "1237511" "5556094" "2332371" ...
##  $ TRANSACTION_TIME  : POSIXct[1:4000000], format: "2020-03-31 18:21:17" "2020-02-07 00:24:34" ...
##  $ TRXN_MONTH        : chr [1:4000000] "3" "2" "6" "3" ...
##  $ CLIENT_ID         : chr [1:4000000] "7116490843" "6249255174" "7117396344" "6249399616" ...
##  $ COUNTRY_NAME      : chr [1:4000000] "United States of America" "India-Republic of" "Switzerland-Swiss Confederation" "United States of America" ...
##  $ COUNTRY_CODE      : chr [1:4000000] "US" "IN" "CH" "US" ...
##  $ CONTINENT_NAME    : chr [1:4000000] "North America" "Asia" "Europe" "North America" ...
##  $ CONTINENT_CODE    : chr [1:4000000] "NN" "AS" "EU" "NN" ...
##  $ SWIFT_MSG_TYPE    : chr [1:4000000] "202" "202" "103" "202" ...
##  $ AVG_TRXN_AMT      : num [1:4000000] 39246 26153 124854 39246 29569 ...
##  $ TRANSACTION_AMOUNT: num [1:4000000] 6475 3335 8920000 1784 2446 ...
##  - attr(*, "spec")=
##   .. cols(
##   ..   TRANSACTION_ID = col_character(),
##   ..   TRANSACTION_TIME = col_datetime(format = ""),
##   ..   TRXN_MONTH = col_character(),
##   ..   CLIENT_ID = col_character(),
##   ..   COUNTRY_NAME = col_character(),
##   ..   COUNTRY_CODE = col_character(),
##   ..   CONTINENT_NAME = col_character(),
##   ..   CONTINENT_CODE = col_character(),
##   ..   SWIFT_MSG_TYPE = col_character(),
##   ..   AVG_TRXN_AMT = col_double(),
##   ..   TRANSACTION_AMOUNT = col_double()
##   .. )

Summary statistics of data frame

summary(input_data_eda)
##  TRANSACTION_ID     TRANSACTION_TIME               TRXN_MONTH       
##  Length:4000000     Min.   :2020-01-01 00:01:48   Length:4000000    
##  Class :character   1st Qu.:2020-03-31 19:41:40   Class :character  
##  Mode  :character   Median :2020-07-03 17:02:13   Mode  :character  
##                     Mean   :2020-07-05 11:03:36                     
##                     3rd Qu.:2020-10-06 00:01:26                     
##                     Max.   :2020-12-31 21:58:20                     
##   CLIENT_ID         COUNTRY_NAME       COUNTRY_CODE       CONTINENT_NAME    
##  Length:4000000     Length:4000000     Length:4000000     Length:4000000    
##  Class :character   Class :character   Class :character   Class :character  
##  Mode  :character   Mode  :character   Mode  :character   Mode  :character  
##                                                                             
##                                                                             
##                                                                             
##  CONTINENT_CODE     SWIFT_MSG_TYPE      AVG_TRXN_AMT    TRANSACTION_AMOUNT 
##  Length:4000000     Length:4000000     Min.   : 20551   Min.   :0.000e+00  
##  Class :character   Class :character   1st Qu.: 26405   1st Qu.:5.438e+03  
##  Mode  :character   Mode  :character   Median : 30873   Median :3.547e+04  
##                                        Mean   : 65288   Mean   :1.156e+07  
##                                        3rd Qu.:105235   3rd Qu.:3.681e+05  
##                                        Max.   :260831   Max.   :1.777e+10

Glimpse of data frame

glimpse(input_data_eda)
## Rows: 4,000,000
## Columns: 11
## $ TRANSACTION_ID     <chr> "3174204", "1237511", "5556094", "2332371", "729...
## $ TRANSACTION_TIME   <dttm> 2020-03-31 18:21:17, 2020-02-07 00:24:34, 2020-...
## $ TRXN_MONTH         <chr> "3", "2", "6", "3", "7", "12", "1", "8", "9", "1...
## $ CLIENT_ID          <chr> "7116490843", "6249255174", "7117396344", "62493...
## $ COUNTRY_NAME       <chr> "United States of America", "India-Republic of",...
## $ COUNTRY_CODE       <chr> "US", "IN", "CH", "US", "US", "US", "RU", "US", ...
## $ CONTINENT_NAME     <chr> "North America", "Asia", "Europe", "North Americ...
## $ CONTINENT_CODE     <chr> "NN", "AS", "EU", "NN", "NN", "NN", "EU", "NN", ...
## $ SWIFT_MSG_TYPE     <chr> "202", "202", "103", "202", "202", "103", "202",...
## $ AVG_TRXN_AMT       <dbl> 39246.11, 26152.55, 124854.38, 39246.11, 29569.2...
## $ TRANSACTION_AMOUNT <dbl> 6475.35, 3335.49, 8920000.00, 1784.00, 2446.45, ...

Head of data frame

head(input_data_eda)
## # A tibble: 6 x 11
##   TRANSACTION_ID TRANSACTION_TIME    TRXN_MONTH CLIENT_ID COUNTRY_NAME
##   <chr>          <dttm>              <chr>      <chr>     <chr>       
## 1 3174204        2020-03-31 18:21:17 3          71164908~ United Stat~
## 2 1237511        2020-02-07 00:24:34 2          62492551~ India-Repub~
## 3 5556094        2020-06-11 13:46:22 6          71173963~ Switzerland~
## 4 2332371        2020-03-10 05:15:07 3          62493996~ United Stat~
## 5 7295929        2020-07-31 17:23:36 7          71164908~ United Stat~
## 6 11840677       2020-12-09 13:52:20 12         71164858~ United Stat~
## # ... with 6 more variables: COUNTRY_CODE <chr>, CONTINENT_NAME <chr>,
## #   CONTINENT_CODE <chr>, SWIFT_MSG_TYPE <chr>, AVG_TRXN_AMT <dbl>,
## #   TRANSACTION_AMOUNT <dbl>

Tail of data frame

tail(input_data_eda)
## # A tibble: 6 x 11
##   TRANSACTION_ID TRANSACTION_TIME    TRXN_MONTH CLIENT_ID COUNTRY_NAME
##   <chr>          <dttm>              <chr>      <chr>     <chr>       
## 1 10019468       2020-10-20 14:40:45 10         71164858~ United Stat~
## 2 2156063        2020-03-04 10:30:43 3          71163786~ Turkey-Repu~
## 3 6491516        2020-07-08 14:32:18 7          71164858~ United Stat~
## 4 1429804        2020-02-13 10:06:57 2          71162836~ Cayman Isla~
## 5 11226381       2020-11-24 06:54:13 11         62493552~ Hong Kong-S~
## 6 4252038        2020-05-01 19:40:36 5          71164908~ United Stat~
## # ... with 6 more variables: COUNTRY_CODE <chr>, CONTINENT_NAME <chr>,
## #   CONTINENT_CODE <chr>, SWIFT_MSG_TYPE <chr>, AVG_TRXN_AMT <dbl>,
## #   TRANSACTION_AMOUNT <dbl>

Segegrate and prepare data for plotting

input_data_plot$AVG_TRXN_AMT=input_data_plot$AVG_TRXN_AMT/1000

input_data_plot <- input_data_plot %>%
  mutate(MONTH_TEXT = case_when(
          (TRXN_MONTH == "1") ~ "Jan",
          (TRXN_MONTH == "2") ~ "Feb",
          (TRXN_MONTH == "3") ~ "Mar",
          (TRXN_MONTH == "4") ~ "Apr",
          (TRXN_MONTH == "5") ~ "May",
          (TRXN_MONTH == "6") ~ "Jun",
          (TRXN_MONTH == "7") ~ "Jul",
          (TRXN_MONTH == "8") ~ "Aug",
          (TRXN_MONTH == "9") ~ "Sep",
          (TRXN_MONTH == "10") ~ "Oct",
          (TRXN_MONTH == "11") ~ "Nov",
          (TRXN_MONTH == "12") ~ "Dec"
    ))

NN_103_df <- input_data_plot[input_data_plot$CONTINENT_CODE =='NN' & input_data_plot$SWIFT_MSG_TYPE=='103',]
NN_103_df <- NN_103_df[,c(3,12,10)]
NN_103_df = NN_103_df %>% distinct()
NN_103_df$TRXN_MONTH = as.integer(NN_103_df$TRXN_MONTH)
NN_103_df <- NN_103_df[order(NN_103_df$TRXN_MONTH),]
glimpse(NN_103_df)
## Rows: 12
## Columns: 3
## $ TRXN_MONTH   <int> 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12
## $ MONTH_TEXT   <chr> "Jan", "Feb", "Mar", "Apr", "May", "Jun", "Jul", "Aug"...
## $ AVG_TRXN_AMT <dbl> 127.9748, 137.2108, 260.8305, 156.2500, 131.9741, 130....
dim(NN_103_df)
## [1] 12  3
NN_202_df <- input_data_plot[input_data_plot$CONTINENT_CODE =='NN' & input_data_plot$SWIFT_MSG_TYPE=='202',]
NN_202_df <- NN_202_df[,c(3,12,10)]
NN_202_df = NN_202_df %>% distinct()
NN_202_df$TRXN_MONTH = as.integer(NN_202_df$TRXN_MONTH)
NN_202_df <- NN_202_df[order(NN_202_df$TRXN_MONTH),]
glimpse(NN_202_df)
## Rows: 12
## Columns: 3
## $ TRXN_MONTH   <int> 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12
## $ MONTH_TEXT   <chr> "Jan", "Feb", "Mar", "Apr", "May", "Jun", "Jul", "Aug"...
## $ AVG_TRXN_AMT <dbl> 29.51093, 24.95118, 39.24611, 32.63196, 28.07124, 31.8...
dim(NN_202_df)
## [1] 12  3
EU_103_df <- input_data_plot[input_data_plot$CONTINENT_CODE =='EU' & input_data_plot$SWIFT_MSG_TYPE=='103',]
EU_103_df <- EU_103_df[,c(3,12,10)]
EU_103_df = EU_103_df %>% distinct()
EU_103_df$TRXN_MONTH = as.integer(EU_103_df$TRXN_MONTH)
EU_103_df <- EU_103_df[order(EU_103_df$TRXN_MONTH),]
glimpse(EU_103_df)
## Rows: 12
## Columns: 3
## $ TRXN_MONTH   <int> 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12
## $ MONTH_TEXT   <chr> "Jan", "Feb", "Mar", "Apr", "May", "Jun", "Jul", "Aug"...
## $ AVG_TRXN_AMT <dbl> 132.44920, 140.76402, 211.85000, 155.32919, 121.98100,...
dim(EU_103_df)
## [1] 12  3
EU_202_df <- input_data_plot[input_data_plot$CONTINENT_CODE =='EU' & input_data_plot$SWIFT_MSG_TYPE=='202',]
EU_202_df <- EU_202_df[,c(3,12,10)]
EU_202_df = EU_202_df %>% distinct()
EU_202_df$TRXN_MONTH = as.integer(EU_202_df$TRXN_MONTH)
EU_202_df <- EU_202_df[order(EU_202_df$TRXN_MONTH),]
head(EU_202_df)
## # A tibble: 6 x 3
##   TRXN_MONTH MONTH_TEXT AVG_TRXN_AMT
##        <int> <chr>             <dbl>
## 1          1 Jan                24.6
## 2          2 Feb                23.0
## 3          3 Mar                24.9
## 4          4 Apr                23.1
## 5          5 May                22.2
## 6          6 Jun                21.8
dim(EU_202_df)
## [1] 12  3
AS_103_df <- input_data_plot[input_data_plot$CONTINENT_CODE =='AS' & input_data_plot$SWIFT_MSG_TYPE=='103',]
AS_103_df <- AS_103_df[,c(3,12,10)]
AS_103_df = AS_103_df %>% distinct()
AS_103_df$TRXN_MONTH = as.integer(AS_103_df$TRXN_MONTH)
AS_103_df <- AS_103_df[order(AS_103_df$TRXN_MONTH),]
glimpse(AS_103_df)
## Rows: 12
## Columns: 3
## $ TRXN_MONTH   <int> 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12
## $ MONTH_TEXT   <chr> "Jan", "Feb", "Mar", "Apr", "May", "Jun", "Jul", "Aug"...
## $ AVG_TRXN_AMT <dbl> 57.26533, 73.94609, 112.92542, 92.02471, 88.45564, 89....
dim(AS_103_df)
## [1] 12  3
AS_202_df <- input_data_plot[input_data_plot$CONTINENT_CODE =='AS' & input_data_plot$SWIFT_MSG_TYPE=='202',]
AS_202_df <- AS_202_df[,c(3,12,10)]
AS_202_df = AS_202_df %>% distinct()
AS_202_df$TRXN_MONTH = as.integer(AS_202_df$TRXN_MONTH)
AS_202_df <- AS_202_df[order(AS_202_df$TRXN_MONTH),]
glimpse(AS_202_df)
## Rows: 12
## Columns: 3
## $ TRXN_MONTH   <int> 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12
## $ MONTH_TEXT   <chr> "Jan", "Feb", "Mar", "Apr", "May", "Jun", "Jul", "Aug"...
## $ AVG_TRXN_AMT <dbl> 29.67334, 26.15255, 23.30047, 25.11515, 25.55450, 24.5...
dim(AS_202_df)
## [1] 12  3

Pie Chart

North America has the most number of wire tranfers.

  library(plotly)

  input_data_plot$pie_count = 1
  
  input_data_plot$CONTINENT_NAME <-  factor(input_data_plot$CONTINENT_NAME, levels=unique(input_data_plot$CONTINENT_NAME))

  plot_ly(input_data_plot, 
          labels = ~CONTINENT_NAME, 
          values = ~pie_count, 
          type = 'pie',
          textposition = 'inside',
          textinfo = 'label+percent',
          insidetextfont = list(color = '#FFFFFF'),
          marker = list(colors = colors,line = list(color = '#FFFFFF', width = 2)),
          showlegend = TRUE)  %>%
  layout(title='<b>Transaction % by Continent</b>',
         xaxis = list(showgrid = FALSE, zeroline = FALSE, showticklabels = FALSE),
         yaxis = list(showgrid = FALSE, zeroline = FALSE, showticklabels = FALSE))
## Warning: `arrange_()` is deprecated as of dplyr 0.7.0.
## Please use `arrange()` instead.
## See vignette('programming') for more help
## This warning is displayed once every 8 hours.
## Call `lifecycle::last_warnings()` to see where this warning was generated.

Bar Plot

These bar plots reveal the average monthly transaction amounts for each cohort of continent, SWIFT message type and month.

library(ggplot2)

options(repr.plot.width = 15, repr.plot.height = 10)

NN_103_df$MONTH_TEXT <-  factor(NN_103_df$MONTH_TEXT, levels=unique(NN_103_df$MONTH_TEXT))
ggplot(NN_103_df) + 
geom_bar( aes(x = MONTH_TEXT, y = AVG_TRXN_AMT), stat = "identity", fill='#5DADE2', color="#000000") +
geom_line(aes(x = MONTH_TEXT, y = AVG_TRXN_AMT), size = 1.5, color="green", group = 1) +
ggtitle("North America MT103 Monthly Median Transaction Amount") +
xlab("Month") +
ylab("Average Transaction Amount (Thousands)") +
theme(axis.text=element_text(size=12),
      axis.title = element_text(size=12),
      plot.title = element_text(hjust = 0.5, size=15,face="bold"))

NN_202_df$MONTH_TEXT <-  factor(NN_202_df$MONTH_TEXT, levels=unique(NN_202_df$MONTH_TEXT))
ggplot(NN_202_df) + 
geom_bar( aes(x = MONTH_TEXT, y = AVG_TRXN_AMT), stat = "identity", fill='#5DADE2', color="#000000") +
geom_line(aes(x = MONTH_TEXT, y = AVG_TRXN_AMT), size = 1.5, color="green", group = 1) +
ggtitle("North America MT202 Monthly Median Transaction Amount") +
xlab("Month") +
ylab("Average Transaction Amount (Thousands)") +
theme(axis.text=element_text(size=12),
      axis.title = element_text(size=12),
      plot.title = element_text(hjust = 0.5, size=15,face="bold"))

EU_103_df$MONTH_TEXT <-  factor(EU_103_df$MONTH_TEXT, levels=unique(EU_103_df$MONTH_TEXT)) 
ggplot(EU_103_df) + 
geom_bar( aes(x = MONTH_TEXT, y = AVG_TRXN_AMT), stat = "identity", fill='#5DADE2', color="#000000") +
geom_line(aes(x = MONTH_TEXT, y = AVG_TRXN_AMT), size = 1.5, color="green", group = 1) +
ggtitle("Europe MT103 Monthly Median Transaction Amount") +
xlab("Month") +
ylab("Average Transaction Amount (Thousands)") +
theme(axis.text=element_text(size=12),
      axis.title = element_text(size=12),
      plot.title = element_text(hjust = 0.5, size=15,face="bold"))

EU_202_df$MONTH_TEXT <-  factor(EU_202_df$MONTH_TEXT, levels=unique(EU_202_df$MONTH_TEXT)) 
ggplot(EU_202_df) + 
geom_bar( aes(x = MONTH_TEXT, y = AVG_TRXN_AMT), stat = "identity", fill='#5DADE2', color="#000000") +
geom_line(aes(x = MONTH_TEXT, y = AVG_TRXN_AMT), size = 1.5, color="green", group = 1) +
ggtitle("Europe MT202 Monthly Median Transaction Amount") +
xlab("Month") +
ylab("Average Transaction Amount (Thousands)") +
theme(axis.text=element_text(size=12),
      axis.title = element_text(size=12),
      plot.title = element_text(hjust = 0.5, size=15,face="bold"))

AS_103_df$MONTH_TEXT <-  factor(AS_103_df$MONTH_TEXT, levels=unique(AS_103_df$MONTH_TEXT))
ggplot(AS_103_df) + 
geom_bar( aes(x = MONTH_TEXT, y = AVG_TRXN_AMT), stat = "identity", fill='#5DADE2', color="#000000") +
geom_line(aes(x = MONTH_TEXT, y = AVG_TRXN_AMT), size = 1.5, color="green", group = 1) +
ggtitle("Asia MT103 Monthly Median Transaction Amount") +
xlab("Month") +
ylab("Average Transaction Amount (Thousands)") +
theme(axis.text=element_text(size=12),
      axis.title = element_text(size=12),
      plot.title = element_text(hjust = 0.5, size=15,face="bold"))

AS_202_df$MONTH_TEXT <-  factor(AS_202_df$MONTH_TEXT, levels=unique(AS_202_df$MONTH_TEXT))
ggplot(AS_202_df) + 
geom_bar( aes(x = MONTH_TEXT, y = AVG_TRXN_AMT), stat = "identity", fill='#5DADE2', color="#000000") +
geom_line(aes(x = MONTH_TEXT, y = AVG_TRXN_AMT), size = 1.5, color="green", group = 1) +
ggtitle("Asia MT202 Monthly Median Transaction Amount") +
xlab("Month") +
ylab("Average Transaction Amount  (Thousands)") +
theme(axis.text=element_text(size=12),
      axis.title = element_text(size=12),
      plot.title = element_text(hjust = 0.5, size=15,face="bold"))

Is the data normally distributed?

library(ggplot2)

input_data_eda <- input_data_eda %>%
  mutate(MONTH_TEXT = case_when(
          (TRXN_MONTH == "1") ~ "Jan",
          (TRXN_MONTH == "2") ~ "Feb",
          (TRXN_MONTH == "3") ~ "Mar",
          (TRXN_MONTH == "4") ~ "Apr",
          (TRXN_MONTH == "5") ~ "May",
          (TRXN_MONTH == "6") ~ "Jun",
          (TRXN_MONTH == "7") ~ "Jul",
          (TRXN_MONTH == "8") ~ "Aug",
          (TRXN_MONTH == "9") ~ "Sep",
          (TRXN_MONTH == "10") ~ "Oct",
          (TRXN_MONTH == "11") ~ "Nov",
          (TRXN_MONTH == "12") ~ "Dec"
    ))


input_data_eda <- input_data_eda[,c(3,12,10)]
input_data_eda = input_data_eda %>% distinct()
input_data_eda$TRXN_MONTH = as.integer(input_data_eda$TRXN_MONTH)
input_data_eda <- input_data_eda[order(input_data_eda$TRXN_MONTH),]
input_data_eda$MONTH_TEXT <-  factor(input_data_eda$MONTH_TEXT, levels=unique(input_data_eda$MONTH_TEXT))
input_data_eda$AVG_TRXN_AMT=input_data_eda$AVG_TRXN_AMT/1000

options(repr.plot.width = 15, repr.plot.height = 10)
ggplot(input_data_eda, aes(x = MONTH_TEXT, y = AVG_TRXN_AMT)) + 
geom_bar(stat = "summary", fun="mean", fill='#5DADE2', color="#000000") +
ggtitle("Monthly Median Transaction Amount") +
xlab("Month") +
ylab("Transaction Amount (Thousands)") +
theme(axis.text=element_text(size=12),
      axis.title = element_text(size=12),
      plot.title = element_text(hjust = 0.5, size=15,face="bold"))

Process Runtime

end_time <- Sys.time()
end_time - start_time
## Time difference of 56.95568 secs